Version 1.6

From Heureka Wiki
Jump to: navigation, search

2011-11-11

Summary

Version 1.6 contains several improvements since last release in beginning of September, for example:

  • We have worked a lot with a function in RegWise to improve handling and computation of the forest state at the start year.
  • Parallelization of the treatment program generator in PlanWise to better utilize the computation capacity of multi-core processors has also been a large activity, in order to decrease computation times.
  • An optimization model for tactical planning problems in PlanWise with road entry decisions has been developed as part of an MSc study for SCA Forest.
  • Forest Domain expressions can be defined as relations between different variables.
  • Dead wood is reported separately for coniferous and deciduous trees.


Important: When using period midpoints in PlanWise, period 0 now refers to a point in time (year 0). Period 1 refers to activities bewteen year 0 to 5, with midpoint 2.5. This can affect your optimization models, see below.

New features

  • The HabitatPrognosis is released (finally!), although we call it a beta version. See download and how-to instructions here. Instructions are in Swedish but the user interface is in English.
  • When editing conditions in a forest domain, expressions can now hold variables with modifiers as expression types. This means that a condition can include a relation between two variables (instead of one variable and a constant value), for example MeanAge >= 1.1*MinFinalFellingAge.
  • RegWise has a new method for calculating forest development to a user-defined start year. The "pre-growth-step" is independent of the settings in the actual scenarios to run. This makes it possible to use exactly the same starting condition for different scenarios when the start year is not the same as the inventory year.
  • Two new thinning guides are available (see control table Treatment Model > Thinning Guide):
    • One is using an up to 6-degree polynomial function to determine what basal area to trigger thinnings, and target basal area after thinning, with dominant height as power variable. The difference with the existing thinning guide (Hugin Thinning Guide), is that the latter approximates a non-linear curve with one or two intersecting straight lines.
    • One is based on stem density. The implementation was financed by Bergvik Skog AB. Three new control table variables are available when selecting this thinning guide:
  • A parameter Min Thinning Grade has been added to control table Treatment Model. It determines the minimum thinning grade accepted in a thinning or selection felling.
  • The default value for Cleaning Period Delay Max in control table Treatment Program Generator (PlanWise and StandWise) has been changed from 2 to 0. This means that the timing of cleaning is not varied, by default, and consequently that more focus is put on varying the thinning schedule and final felling period. We have found little point in varying the timing of cleanings with respect to predicted growth and yield, the difference seems marginal. Still, the net present value can indeed be affected by the timing of the cleaning, so if the purpose of an analysis is to study the optimal time for cleaning this setting should of course be modified.
  • When editing a report template, the user can now choose a default view for the report (graph, table, or mixed).
  • In control tables, a single variable can be reset to its default value (right-click and select "Reset to default value"). Before, you could only reset the entire table.
  • Volume harvested is computed also in cleanings.
  • Handling of root extraction for bio fuel has been improved, both with respect to cost calculations and possibility to select for which species that roots and stumps can be harvested (variable Stump Extraction in control table Treatment Model). Only spruce stumps are harvested by default. Instead of a fixed cost, a linear function of terrain transport distance is used, and different functions are used for thinning and final felling, and in final felling, for harvest residues and stump.
  • Strip road area is proportional to tree spacing. If spacing between trees is larger than strip road width, there will be no strip road area.
  • Volume of dead wood is now reported as "Downed deadwood", and "Standing Dead Trees" for, yes, volume of lying and standing dead wood. Also, dead wood is reported separately for coniferous and deciduous trees.

Performance improvements

  • The treatment program generator in PlanWise is faster on multi-core processor PCs (which most PC:s are today). This has been achieved through parallellization of the program so that the computer's capacity is better utilized.
  • The evaluation of forest domains when running Tactical Treatment Program Generator has been very time consuming, but this has been fixed.

Bug fixes

  • The number of selection fellings was too few under certain conditions when applying uneven-aged management (CCF).
  • When calculating initial tree age a function is used to determine if a tree should be considered an overstorey tree. If so, a specific age function is used. For different reasons, this functionality had been removed, but is now re-implemented.
  • For ingrown trees, the function for uneven-aged stands is now used to compute tree age, independent of whether the stand is classified as even-aged or uneven-aged. As a consequence of this fix, ingrown trees get smaller ages, which in turn gives a larger growth than before. In an example that we ran for a spruce stand with H100 = 33 m and a mean age of 68 years, a certain ingrown tree is now 29 years, compared to 59 years before the change.
  • When adding a new "user defined integer set" to an optimization model, an error occured in the index for existing parameters due to a naming conflict. This has been corrected, so that a unique index is generated.
  • In the Timber Pricelist editor (control table Pricelist, section Prices > Timber), there is a setting "Pulpwood proportion" for how many of the butt logs, middle logs, and top logs, respectively, that are downgraded to pulp. Probably, few users have changed these settings which are 10 % by default. However, changing these values have not lead not to expected results, because the value for top log and but log had been switched. This has been fixed. Also, the focus is always on the first tab (Pine) when opening the editor.

More bug fix reports are available here.

Other stuff you should know

  • When using period midpoints in PlanWise strategic TPG, period 0 now refers to year 0, not year 2.5. Instead, the period midpoint for period 1 is year 2.5, and so forth. Hence, period 0 is not an actual period, but a point in time, followed by a series of 5-year periods. The problem have been that treatments applied in year 0, for example regeneration in recently clear cut stands, were not reported when using period midpoint, but now they are. Note that no regular harvesting will take place in period 0, only removal of existing seed trees and shelterwoods. This will probably affect existing optimization models, if you have restrictions that include harvesting in period 0. Add condition "with p > 0" when declaring such constraints, for example
forall <p> in periods with p > 0 do


Next release

  • Date is not decided...
  • A new stand register version (2) will be released (actually PlanStart supports it but we have not documented it yet). The new version has columns for initial dead wood volume. The previous version (1) will still be supported.